MERGE TO GYRENE.

551 explore this passage to the end. It is formed entirely in the rock from which the stream issues, and runs, in an irregular course, for nearly a quarter of a mile into the bowels of the mountain : the sides and roof of the passage are flat, where time and the action of the current (which is very strong) have not worn them away ; but the bottom is encumbered with stones, bedded fast in a quantity of clay which has accumulated about it and against the sides. The general height of this subterranean channel is scarcely five feet, an elevation which we found rather inconvenient, for it obliged us to stoop a good deal in advancing ; and as it would not have been possible to examine the place properly, or indeed to have preserved our light, without keeping the head and body in an upright position, we usually found the water making higher encroaches than its chilling cold rendered agreeable.

In some places, however, where there appear to have been originally flaws or fissures in the rock, the roof was irregular, and there was room enough to stand upright, an occurrence of which we very gladly availed ourselves, to the great relief of our knees. We found the average width from three to four feet, although in the places just mentioned it was occasionally as much as six feet ; and were it not for the clay which has been collected against the sides, we should often have suffered from their roughness. F rom the irregularity of the course of the passage we were obliged to take bearings very often ; and at each time w'e stopped for this purpose we took down the distance measiu-ed with our chain between the point we stopped at and the last ; so that after much trouble we succeeded in obtaining a tolerably correct plan of the